Robuste techniques for multivariate surrogate modeling

01 October 2008 → 30 September 2014
Regional and community funding: Special Research Fund, Research Foundation - Flanders (FWO)
Research disciplines
  • Engineering and technology
    • Electronics
    • Modelling
    • Multimedia processing
multivariate macromodel modeling
Project description

The computation of accurate and compact surrogate models, which represent the broadband input-output behavior of complex linear systems, has become a topic of intense research, mainly due to the increasing speed and decreasing size of electronic circuits. The objective of this research project is to develop algorithms which are able to generate guaranteed passive, rational, broadband circuit models, based on simulated time domein or frequency domain data. This data may depend on multiple physical design variables (like the lenght, width, or material parameters of the structure), which leads to a multidimensional approximation problem.